The present invention relates generally to an image pickup system and an image pickup optical system, and more particularly to an image pickup optical system and electronic image pickup system for forming a subject image on the image pickup surface of an electronic image pickup device such as a CCD, which comprises a plurality of pixels having at least three different spectral characteristics for obtaining a color image.
An electronic image pickup system using an electronic image pickup device is designed to have sensitivity even to wavelengths of 400 nm or less so as to ensure the quantity of light sensible to the electronic image pickup device. For instance, when the quantity of light sensed by the electronic image pickup device is small, gamma characteristics are often controlled to make the output from the photoelectric converters higher than the input thereto upon image reproduction. If, in this case, the spectral state of the subject is substantially constant, no particular problem arises. However, when the energy of wavelengths (e.g., h-line) in the vicinity of 400 nm is large with respect to that of, for instance, 450 nm or g-line, a problem arises; for instance, the blue tint of the reproduced image is more stressed than when actually seen by the human eyes. The reason is that while the sensitivity of the human eyes to the short wavelength side of the visible range is considerably low, yet such short wavelengths are reproduced by an electronic image pickup device in colors perceptible to the human eyes, because the sensitivity of the device to the short wavelength range is relatively high.
For recently developed digital cameras comprising a ever-larger number of pixels, on the other hand, it is required to achieve drastic size and cost reductions. With this, image pickup optical systems, too, are now required to have ever-higher performance and ever-higher zooming and other functions with size and cost reductions. To achieve higher performance, it is required to increase the image-formation capability of a system all over the wavelength range to which the system is sensible. In the present disclosure, changes in the image-formation capability due to wavelengths are called chromatic aberrations. In general, the chromatic aberrations are corrected making use of the fact that the rate (dispersion) of change of the index of refraction with respect to wavelengths differ from material to material. For instance, an optical system having a positive focal length is designed to make correction for the chromatic aberrations by using a material of small dispersion for an optical element having positive refracting power and a material of large dispersion for an optical element having negative refracting power.
When the chromatic aberrations are corrected by using a combination of optical elements as mentioned above, it is required to take not only the chromatic aberrations but also the image-formation capability of the whole image pickup surface into consideration; for instance, it is required to increase the number of optical elements. For a zoom lens system of the type that the focal length of the system is varied by varying the separations between a plurality of lens groups comprising a lens group having a positive focal length and a lens group having a negative focal length, more complicated combinations of optical elements are required. In this case, when a refracting type of optical element (lens) is formed using a glass or plastic material, the index of refraction increases, sometimes drastically, as the wavelength changes from a long wavelength side to a short wavelength side, although depending on the material used.
FIG. 84 is illustrative of how the refractive index of two single lenses whose refracting power (the reciprocal of the focal length) becomes 1 at 550-nm wavelength change due to wavelengths. It is here noted that the single lenses are constructed using a typical vitreous material and a material called an ultra-low dispersion material. FIG. 85 is illustrative of the amount of displacements on the basis of 500 nm of the back focal position of an optical system constructed only of a general refraction type optical element, with wavelength as abscissa and displacement as ordinate. As can be seen from FIG. 84, the refracting type optical element shows a similar power change tendency with respect to wavelength, irrespective of whether it is formed of a normal material or an ultra-low dispersion material. Thus, the axial chromatic aberration of an image pickup optical system constructed of a refracting optical element formed of a material in the practical range has a V-shaped form as shown in FIG. 85; an image is formed on the same point at only two wavelengths with the chromatic aberrations becoming large both on the short and long wavelength sides. On the short wavelength side in particular, there is a large chromatic aberration change. To reduce such chromatic aberration changes, it is proposed to make use of fluorite and a special-purpose glass such as an ultra-low dispersion glass. However, this special glass, too, has such characteristics as shown in FIG. 84; in other words, it is difficult to reduce the chromatic aberration change on the short wavelength side to a sufficiently low level. When this glass is used for an electronic image pickup device, colors of short wavelengths appear together with chromatic aberrations, and so offer an unnatural “color running” problem.
In JP-A 10-170822 as one prior art, it is proposed to reduce the chromatic aberration changes on the short wavelength side by use of a diffractive optical element. According to this publication, chromatic aberrations due to the light used are corrected making use of the reciprocal dispersion of the diffractive optical element. In the diffractive optical element, however, diffracted light other than the used light appears in the form of unnecessary light, which is in turn responsible for ghosts and flares. In this publication, the wavelength range is limited, whereby the influence of unnecessary light on the diffractive optical element is reduced.
However, this unnecessary light reaches the image-formation plane discontinuously (independently) with respect to the used light. In the spectral wavelength characteristics, too, the unnecessary light is discontinuous with respect to the used light. In this publication, unnecessary light having no relation to normal image-formation (by the used light) is reduced making use of its wavelength characteristic difference. To obtain good images, it is thus required to largely reduce the strength of the image due to unnecessary light. For instance, a system hardly sensible to 420 nm is proposed.
However, 420 nm has an influence on the sense of sight of the human in general and the perception of colors in particular. In view of color reproduction, reducing this wavelength is tantamount to reducing a short wavelength component than required, leading to a possible impairment of natural color reproduction. Thus, a problem with the technique set forth in this publication is that it is difficult to make a reasonable tradeoff between high color reproducibility and flare removal, because the short wavelength range having an influence on the sense of sight of the human must be largely cut off so as to make the influence of unnecessary light unobtrusive.
With conventional design with weight given to an intermediate wavelength region in the visible range, it is impossible to make perfect correction for chromatic aberrations at both ends of the visible range, and those on the short wavelength side in particular. For this reason, when the image of a high-contrast subject is picked up, the colors of shorter wavelengths are not only stressed but also color flares of brighter blues occur at the boundary of light and shades.
For recently developed digital cameras comprising a ever-larger number of pixels, it is required to achieve drastic size and cost reductions, as already mentioned. With this, image pickup optical systems, too, are now required to have ever-higher performance and ever-higher zooming and other functions with size and cost reductions. Especially for increasing the number of pixels and achieving size reductions, it is required to decrease the area of each pixel of the image pickup device. This means that it is required to increase on a per-unit-area basis the quantity of light subjected to photoelectric conversion by an image pickup device. In other words, it is required to make the S/N ratio of the device favorable, maintain the sensitivity of the device to a dark subject and make short device exposure time To obtain a color image, a color filter having such a filter arrangement as shown in FIG. 2 or 3 is located in front of the image pickup device so as to achieve a photoelectric conversion device having at least three different wavelength characteristics. The filter shown in FIG. 2 is of the type called a primary color filter comprising red (R), green (G) and blue (B) filter elements. The respective wavelength characteristics of these filter elements are shown in FIG. 4. The filter shown in FIG. 3 is of the type called a complementary color filter comprising cyan (C), magenta (M), yellow (Ye) and green (G) filter elements. The respective wavelength characteristics of these filter elements are shown in FIG. 5. When the complementary color filter is used as the filter, the filtered light is converted by a controller 4 to R, G and B according to the following processing:
for luminance signalsY=|G+M+Ye+C|*¼for color signalsR−Y=|(M+Ye)−(G+C)|B−Y=|(M+C)−(G+Ye)|
Both the primary color filter and the complementary color filter are not sensible to the human eyes. In many cases, an IR cutoff filter having sensitivity to the image pickup device and capable of cutting off light of wavelengths of about 700 nm or greater (infrared cutoff filter) is located in an optical system. Most of IR cutoff filters are designed to cut off wavelengths of 700 nm or greater, and so their transmittance with respect to the vicinity of 600 nm becomes worse, as shown in FIG. 68.
With the primary color filter, it is easy to carry out processing for color reproduction. When the complementary color filter with an R, G and B conversion process is used, an output of R signals (for red development) is produced with respect to the input of the blue wavelength range (a wavelength of about 400 nm to about 430 nm in FIG. 11) upon conversion from the complementary color filter to R, G and B.
For this reason, the primary color filter is mainly used for digital cameras required to have a large number of pixels and high image quality. Sometimes, the complementary color filter is used for an image pickup system less likely to produce chromatic aberrations. The “image pickup system less likely to produce chromatic aberrations” is understood to include an image pickup system wherein the number of pixels is so reduced that the aberrations of a phototaking lens have little or no influence on image quality, and an image pickup system with inherently reduced chromatic aberrations (this may be achieved by increasing the F-number of the system, decreasing the magnification of a zoom lens in the case of a zoom lens system, using a vitreous material (e.g., fluorite) that costs much or is of poor productivity, increasing the number of lens elements, and increasing the length of the system.
With the primary color filter, it is easy to carry out processing for color reproduction. However, the quantity of light entering each pixel is small (because the wavelength range of light entering each pixel is narrow). In the primary color filter, only G has sensitivity to green (light in the wavelength range of about 500 nm to about 550 nm) that is a significant determinant for image resolution. For this reason, the primary color filter is designed in such a way that the ratio of R, G and B pixels is set at 1:2:1, thereby regulating the ratio of pixels having a significant influence on the determination of image resolution to 50%.
When such a primary color filter is used, the quantity of light entering each pixel is small, and so problems arise in connection with S/N and exposure time as pixel size becomes small. The pixels having an influence on image resolution are barely 50%; there is a problem that it is impossible to take full advantage of the large number of pixels, thereby achieving high image quality.